Fig 1.
Landscape of sox genes in representative vertebrate genomes.
The sox genes (top) are divided into the 7 groups: B1, B2, C, D, E, F and H. Phylogenetic relationships of the different analyzed vertebrate species are indicated on the left. The orange and red circles on the phylogeny represent the teleost-specific WGD and the salmonid-specific WGD, respectively. Light blue squares indicate gene singletons. Orange and dark blue squares indicate duplicates produced either by the teleost-specific WGD or by small-scale duplications (SSDs) respectively. Red squares correspond to genes detected in multiple copies (two, three or four as indicated by the number in the square) in salmonids. White squares are used when no copy was detected. The mammal-specific SoxA group is not represented on the figure.
Fig 2.
Macrosynteny analyses of duplicated sox genes in teleosts.
(a-d) All sox duplicates annotated in the four zebrafish, medaka, stickleback and Tetraodon genomes were studied in the macrosynteny analysis using the last release of genome assemblies in Ensembl. Grey lines connect paralog genes on the different chromosomes or linkage group in the genomes. Orange lines connect paralog sox genes on different chromosomes.
Fig 3.
Phylogeny analyses of duplicated sox genes in teleosts.
(a-i) Phylogenetic reconstructions were done for the nine genes sox1, sox4, sox6, sox8, sox9, sox10, sox11, sox14 and sox21. Teleost-specific whole genome duplication paralogs are named soxa in blue and soxb in orange. Non-teleost orthologs are represented in grey. The most closely related human SOX gene was used to root the tree. Significant asymmetric evolution between sox paralogs in teleosts is highlighted on the phylogeny (see S4 Fig for details) using: NS for non significant, * for significant (P < 5%), ** for highly significant (P < 1%), and *** for extremely significant (P < 0.1%). Phylogenies were computed using PhyML and based on complete protein sequences alignment from mouse M. musculus (Mm), human H. sapiens (Hs), chicken G. gallus (Gg), coelacanth L. chalumnae (Lc), spotted gar L. oculatus (Lo), zebrafish D. rerio (Dr), catfish C. gariepinus (Cg), mud loach M. anguillicaudatus (Ma), salmon S. salar (Ss), cod G. morhua (Gm), tilapia O. niloticus (On), medaka O. latipes (Ol), platyfish X. maculatus (Xm), guppy P. reticulata (Pr), tetraodon T. nigroviridis (Tn), fugu T. rubripes (Tr) and stickleback G. aculeatus (Ga).
Fig 4.
Conserved non-coding elements associated to sox4, sox8, sox9, sox10 and sox11.
Conserved non-coding elements (CNEs) are highlighted in the vicinity of fish sox genes. The spotted gar has been used as a reference. Blue and orange circles represent CNEs in the vicinity of WGD-derived paralogs a and b of sox4 (a), sox8 (b), sox9 (c), sox10 (d) and sox11 (e). Grey circles are used for spotted gar, human and mouse CNEs. (f) Numbered representation of panels a to e above. Sox11b of the stickleback is not included in this study as there are no enough sequences available around the gene.
Fig 5.
Expression patterns of sox4, sox8, sox9, sox10 and sox11 in three teleost species.
(a) Expression patterns during D. rerio (Dr) and O. latipes (Ol) embryonic development. The five major developmental periods segmentation, gastrula, neurula, pharyngula and hatching are indicated on the left. (b) Expression patterns in seven adult tissues of D. rerio (Dr), O. latipes (Ol), X. maculatus (Xm) and M. musculus (Mm). qRT-PCR experiments were performed during embryonic development and adulthood. Data were normalized with the two housekeeping genes bActin2 and rpl7 and analyzed by the ΔΔCT method (Livak and Schmittgen 2001). Low expressed genes are indicated in grey (ΔΔCT≤ 0.10). More expressed genes are indicated in blue, the intensity of the blue increasing with the intensity of expression as indicated on the figure.